EP2510391B1 - Device for coupling an electromagnetic wave between a wave guide and a grooved metal guide, method for manufacturing said device, and optical electric coupler for an object using the optical coupling device - Google Patents

Description

Technical field of the invention

• un guide d'onde,
• un guide métallique à fente.

The invention relates to a device for coupling an electromagnetic wave comprising:

• a waveguide,
• a slotted metal guide.

State of the art

The size of the electronic components is constantly decreasing, allowing the increase of the integration density of these components on the same chip.

In the field of optoelectronics, it has been necessary to reduce the size of the waveguides accordingly to optically access these components with good efficiency. Without particular precautions, the coupling efficiency is substantially equal to the ratio between the optical mode surface carried by the waveguide and the characteristic surface of the device to be addressed.

The miniaturization of optical circuits has already led to a miniaturization of planar optical guides, whose section can go down to dimensions of the order of 200 nm by 400 nm. However, these decreases are not sufficient when trying to access, for example, a 50nm nanowire.

The document "Theorical investigation of compact couplers between dielectric-wave waveguides and two-dimensional metal-dielectric-metal plasmonic waveguides" published in "Optic Express" Vol. 15, No. 3 of February 5, 2007 describes a device for coupling an electromagnetic wave. Such a device, illustrated in Figures 1, 2 and 3 uses a silicon guide in two elements 1a, 1b and a silver guide 2 of metal-dielectric-metal type interposed between the two elements 1a, 1b. The elements 1a, 1b and the silver guide 2 are arranged end-to-end. The metal-dielectric-metal type guide 2 consists of two silver elements 3a, 3b separated by a slot 4 filled with dielectric. The slot 4 thus delimited is substantially coaxial with the central axis of the elements forming the silicon guide 1 a, 1 b, where the amplitude of the electromagnetic field generated by the wave is maximum. Such a device is complex to manufacture because it requires a perfect fit of the different elements put end-to-end. This structure forms a resonant device whose transmission rate is very sensitive to the incident wavelength. In addition, such a device induces a reflection problem, in fact, a portion of the incident wave moves in the silicon guide 1a and is transmitted in the silver guide 2 but another is reflected. Thus, such a coupling is not satisfactory because it induces a significant decrease in the efficiency of the coupling of the electromagnetic wave.

Object of the invention

An object of the invention is to provide a compact and efficient coupling between a waveguide and a slot metal guide.

This object is attained by the appended claims and more particularly by the fact that the slotted metal guide is formed by two coplanar metal elements spaced from one another so as to define the slot, and in that the guide slit metal is disposed in a plane offset from the plane of the waveguide and partially overlaps said waveguide, said waveguide and slot guide being kept at a distance from each other by a dielectric.

The invention also relates to a method of manufacturing a coupling device characterized by claims 11-15.

The invention also relates to a claimed optical and electrical coupler of an object, said coupler comprises an optical coupling device and an electrical coupling element provided with two electrodes arranged at the slot of the slot guide, said object being in contact with at least one of the electrodes.

Brief description of the drawings

• La figure 1 illustre un dispositif de couplage selon l'art antérieur.
• La figure 2 illustre le dispositif de couplage de la figure 1 selon une coupe A-A.
• La figure 3 illustre le dispositif de couplage de la figure 1 selon une coupe C-C.
• La figure 4 illustre respectivement la propagation d'une composante TM et TE d'une onde électromagnétique dans un guide d'onde.
• La figure 5 illustre une vue de dessus d'un mode de réalisation selon l'invention.
• La figure 6 illustre une vue en coupe selon A-A de la figure 5.
• La figure 7 illustre un autre mode de réalisation d'un dispositif selon l'invention.
• La figure 8 illustre des courbes représentatives de l'évolution de l'indice effectif en fonction de la largeur d'un guide d'onde en silicium.
• La figure 9 illustre un autre mode de réalisation de l'invention.
• La figure 10 illustre une vue de côté du dispositif de la figure 9.
• La figure 11 illustre un autre mode de réalisation de l'invention.
• Les figures 12 à 16 illustrent un procédé de fabrication d'un dispositif selon l'invention.
• Les figures 17 à 19 illustrent deux modes de réalisation d'un coupleur optique et électrique d'un objet.

Other advantages and features will emerge more clearly from the following description of particular embodiments of the invention given by way of non-limiting example and represented in the accompanying drawings, in which:

• The figure 1 illustrates a coupling device according to the prior art.
• The figure 2 illustrates the coupling device of the figure 1 according to a AA cut.
• The figure 3 illustrates the coupling device of the figure 1 according to a CC section.
• The figure 4 illustrates respectively the propagation of a TM and TE component of an electromagnetic wave in a waveguide.
• The figure 5 illustrates a top view of an embodiment according to the invention.
• The figure 6 illustrates a sectional view according to AA of the figure 5 .
• The figure 7 illustrates another embodiment of a device according to the invention.
• The figure 8 illustrates representative curves of the evolution of the effective index as a function of the width of a silicon waveguide.
• The figure 9 illustrates another embodiment of the invention.
• The figure 10 illustrates a side view of the device of the figure 9 .
• The figure 11 illustrates another embodiment of the invention.
• The Figures 12 to 16 illustrate a method of manufacturing a device according to the invention.
• The Figures 17 to 19 illustrate two embodiments of an optical and electrical coupler of an object.

Description of preferred embodiments

As illustrated in figure 4 when an electromagnetic wave also called in this optical wave domain (for a wavelength range of visible to near and intermediate infrared) propagates along a longitudinal axis of the waveguide 1 as indicated by the vector of wave k, this wave breaks down into two states of polarization. A first state forms a transverse electrical component (TE "Transverse Electric" in English) and a second state forms a transverse magnetic component (TM for "Transverse Magnetic"). The TM component comprises an electric field E perpendicular to the plane of the guide and a magnetic field H parallel to the plane of the guide. The TE component comprises a magnetic field H perpendicular to the plane of the guide and an electric field E parallel to the plane of the guide.

As illustrated in Figures 5 and 6 , the coupling device of an electromagnetic wave (that is to say for example an optical wave as described above) comprises a waveguide 1, for example formed by a silicon guide, which may be under the shape of a ribbon. Such a waveguide may be a type III or IV material or an alloy of these types of materials such as InP, GaS, InGaS. Advantageously, the waveguide 1 is transparent to the wave considered. In general, the waveguide 1 may be a partially engraved (guide stops) and / or locally structured planar guide for containing, for example, networks for coupling the waveguide 1 to an optical fiber. The coupling device further comprises a slotted metal guide 2, preferably formed by two metallic elements 3a, 3b coplanar and spaced from each other so as to form said slot 4. By way of example, the elements metal 3a, 3b may be formed of silver or copper.

The slotted metal guide 2 is disposed in a plane P2 offset from the plane P1 of the waveguide 1 and partially covers said waveguide 1 ( figure 6 ). This overlap allows the interaction of the evanescent fields of the two guides to achieve the coupling. Preferably, the two guides 1, 2 are kept at a distance from one another by a dielectric for limiting the reflection, diffusion or absorption of the wave. In other words, the waveguide 1 and the slot metal guide 2 can be separated by an interval D _i . The dielectric may be in the form of a layer on the one hand in contact with the waveguide 1, and on the other hand in contact with the slotted metal guide 2, it can therefore have a thickness equal to Di. This dielectric may be a silicon oxide layer. More generally, the dielectric material used may have a low refractive index with respect to the refractive index of the silicon guide. Preferably, this refractive index is between 1 and 2.5, and best confers the optical wave in the waveguide 1 to prevent the latter from diffusing outside said waveguide 1. The waveguide may have, by way of example, an index of the order of 3.55.

When an electromagnetic wave propagates in a guide, a part of this wave penetrates into the surrounding medium, thus generating a so-called evanescent field. This length of penetration is proportional to the ratio of refractive index of the guide and its surrounding environment. By coupling a waveguide 1 with a slotted metal guide 2 as described, the evanescent field of the waveguide can be progressively accumulated by the slotted metal guide 2 when the two guides are close to one another. the other. This process is also reciprocal.

In an embodiment illustrated in Figures 5 and 6 , the waveguide 1 is constituted by a single element forming a ribbon, preferably made of silicon. The metal slotted guide 2, for example made of silver, is formed by two metallic elements 3a, 3b, coplanar, and spaced from each other so as to delimit said slot 4. The slot 4 is preferably oriented along a longitudinal axis A1 of the waveguide 1. Preferably, the slot 4 is centered so that it lies in a plane perpendicular to the plane of the waveguide 1 and passing through its center, the intersection of the two planes then forms the longitudinal axis A1 of the waveguide 1. The view of the figure 5 being a top view of the coupling device, the slotted metal guide 2 is disposed above the waveguide 1 so that the slotted metal guide 2 partially covers the waveguide 1. The embodiment of FIG. Figures 5 and 6 and is not limiting, so in general, the slotted metal guide 2 and the waveguide 1 can be substantially disoriented and off-center.

According to an embodiment variant illustrated in figure 7 , the waveguide 1 consists of two elements 1a and 1b which are not made on the same axis, the slot metal guide 2 is then bent to connect the two elements 1a and 1b. On the figure 7 , the two elements 1a and 1b are spaced from each other and their longitudinal axes are perpendicular.

According to one particular embodiment, the electromagnetic field in the slotted guide being transverse, only the eigenmodes of the polarized waveguide 1 TE are coupled, that is to say that only the TE component propagates in the guide to slot 2 while the TM component remains confined to the waveguide 1. Thus, the coupling device is also called co-directional coupler.

In the exemplary embodiment based on silicon waveguide 1 and silver slot 2 metal guide, the coupling is performed between the fundamental modes (of higher effective index) TE of the two guides, this mode coupled to slotted metal guide 2 allows a confinement of the coupling in the slot of said slot guide 2. The power density obtained in the slot guide 2 is then exacerbated of the order of 100.

The use of evanescent fields makes it possible to increase the efficiency of the coupling with respect to the end-to-end coupling of the prior art. The coupling mechanism between the waveguide 1 and the slot metal guide 2 can be illustrated using equations obtained in a perturbative approach of the Maxwell equations, for a simplified system where the two guides are identical. These equations will be sufficient to translate the propagation of the wave in a coupled system and the parameters that influence this coupling. A thorough description is made by A. Yariv in "coupled mode theory for guided wave optics", IEEEJ. As. Elec. 9, 919, 1973.

The waveguide 1 and the slot metal guide 2 may be single-mode or multi-mode depending on the width and / or the height of the waveguide 1 and the width and / or thickness (or height) of the metallic elements 3a, 3b of the slot metal guide 2. It is also possible to play over the width of the slot to make a selection of mode (s) to be coupled and thus form a mode filter (s) . According to one development, a mode of the waveguide 1 and a mode of the slotted metal guide 2 have the same effective index and are coupled. A mode corresponds to a spatial configuration of the electromagnetic field induced by the propagation of an electromagnetic wave in the guide associated with said mode. According to the distribution of the electromagnetic field, a given mode sees a different index of the materials constituting the guide and the index seen by each mode defines an effective index. An effective index can be defined as the ratio of the wavelength in the vacuum to the wavelength in the medium to be crossed.

avec $$L_c=\frac{\lambda }{2\sqrt{{\mathrm{\Delta }\mathit{n}}_{\mathit{eff},S}^2+{\mathrm{\Delta }\mathit{n}}_{\mathit{eff},L}^2}}$$

• où, λ est la longueur d'onde d'opération (onde se déplaçant dans le dispositif de couplage),
• Δn_{eff, L} est la différence entre les indices effectifs des modes supportés par les guides en l'absence de couplage. Ces modes sont appelés modes locaux.
• Δn_{eff, S} : est la différence entre les indices effectifs des modes (appelés supermodes) supportés par la structure de couplage comprenant les deux guides juxtaposés (guide d'onde et guide métallique à fente). Ces deux supermodes sont par construction de symétrie opposée. On parle alors de supermodes symétrique et antisymétrique. Cette quantité physique est dépendante de l'intervalle Di qui sépare les deux guides. Cette distance Di conditionne la longueur de couplage L_c ainsi que le taux de puissance transférée.

The optimum coupling efficiency ( F ) is defined as the fraction of power transmitted from a first guide to a second guide after a characteristic distance L _c also called effective coupling length: $$F=\frac{1}{1+{\left(\frac{{\mathrm{\Delta }\mathit{not}}_{\mathit{eff},\mathrm{The}}}{{\mathrm{\Delta }\mathit{not}}_{\mathit{eff},\mathrm{<figure-callout\; id="2"\; label="S"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">S</figure-callout>}}}\right)}^2}$$

with $${\mathrm{The}}_{\mathrm{vs}}=\frac{\mathrm{<figure-callout\; id="2"\; label="\lambda "\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">\lambda </figure-callout>}}{2\sqrt{{\mathrm{\Delta }\mathit{not}}_{\mathit{eff},\mathrm{<figure-callout\; id="2"\; label="S"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">S</figure-callout>}}^2+{\mathrm{\Delta }\mathit{not}}_{\mathit{eff},\mathrm{The}}^2}}$$

• where, λ is the operating wavelength (wave moving in the coupling device),
• Δ n _{eff, L} is the difference between the effective indices of the modes supported by the guides in the absence of coupling. These modes are called local modes.
• Δ n _{eff, S} : is the difference between the effective mode indices (called supermodes) supported by the coupling structure comprising the two juxtaposed guides (waveguide and slot metal guide). These two supermodes are by construction of opposite symmetry. This is called symmetric and antisymmetric supermodes. This physical quantity is dependent on the interval Di which separates the two guides. This distance Di conditions the coupling length L _{c as} well as the transferred power rate.

The transfer of power from one guide to another is a harmonic process and therefore reversible. Considering a nominal power P ₀ injected into the waveguide 1, the power collected or transmitted in the slot metal guide 2 at the position z is expressed as follows: $${\mathrm{<figure-callout\; id="1"\; label="P"\; filenames="imgf0002.png,imgf0007.png"\; state="\{\{state\}\}">P</figure-callout>}}_{1\to 2}\left(z\right)=P_0\cdot \mathit{<figure-callout\; id="2"\; label="F\; sin"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">F\; </figure-callout>}{\sin }^{}{}^2\left(\frac{\mathrm{<figure-callout\; id="2"\; label="\pi "\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">\pi </figure-callout>}}{2{\mathrm{The}}_{\mathrm{vs}}}\cdot z\right)$$

According to relation (3), the power transfer from the first guide to the second guide is optimal for odd multiples of the coupling characteristic length L _c .

The efficiency of the coupling can be improved as a function of the distance separating the waveguide 1 from the slotted metal guide 2, their effective indices, the length of the slotted metal guide 2 or the partial overlap length between the guides.

Thus, according to a first improvement, it is preferable to use two guides 1, 2 whose effective indices n _eff , uncoupled modes, are substantially equal so that the term F of equation (1) tends towards the unit. Preferably, the ratio of the effective index of the waveguide 1 with the effective index of the slotted metal guide 2 is between 0.8 and 1.2, and preferably equal to 1. The effective index of a mode can vary according to the dimensions (width and thickness) of the guide considered, such a variation is described in "Silicon photonics" Reed, GT, Knights, AP, ed. Wiley (2004 ). In order to allow an optimal interaction of the modes of the waveguide 1 and the slot metal guide 2, it is possible to fix the effective index of a first guide, then to adjust the geometric dimensions of a second guide to couple with the first to obtain an effective index agreement between the eigen modes of each of the guides.

In the example, the effective index of the modes supported by these guides can be calculated numerically by a mode solver by using the FDTD method (for finite-difference frequency-domain in English). In this particular example, the geometry, and thus the effective index n _eff , of the slot guide 2 is set. The dimensions of the silicon waveguide 1 will be characterized by fixing, for example, its height, and by varying its width to obtain an effective index substantially equal to that of the slot metal guide 2. The effective index of the guide slit metal 2 depends on the width l ₁ of the slot 4, the height h ₁ of the slot 4 ( figure 6 ) and the type of material used. According to the example, the slot metal guide 2 is made of silver, the width l ₁ of the slot 4 is 25 nm and the height h ₁ of the slot 4 is 50 nm, the effective index of such a guide is set at 2.25 at the figure 8 (Ag curve). On this figure 8 , a curve Si is also reported, this curve illustrates the evolution of the effective index n _eff of the silicon guide with respect to its width Wsi. Therefore, the figure 8 allows to determine a point of intersection of the curves Ag and Si where the effective indices of the two guides are equal. So, for the silver slot metal guide 2 previously described, the preferential width of the associated silicon guide is 410 nm.

In this particular example, the width of the silicon waveguide 1 thus depends on the effective index of the slot metal guide 2. This width of the silicon guide 1 will be different if the metal of the slot guide is changed or if the dimensions of the slot 4 vary.

According to a second improvement, the interval Di separating the two guides 1, 2 is adjusted to optimize the length and the coupling efficiency. In the particular case where the effective indices of the two guides when the latter are not coupled are strictly identical, equation (1) is equal to one and the interval Di affects only the coupling length Lc and therefore the compactness of the device.

According to the particular example using a silicon ribbon guide 1 and a slotted silver guide 2, four test devices each associated with a separation interval de of 250nm, 150nm, 100nm, and 50nm were realized. The graph of the figure 8 illustrates the variations of the effective indices of the even and odd supermodes of the structure as a function of the width of the silicon guide. The curves C1, C2, C3, C4 are respectively associated with the distances 250nm, 150nm, 100nm, and 50nm in the odd supermode and the curves C1 ', C2', C3 ', C4' are respectively associated with the distances 250nm, 150nm, 100nm , and 50nm in the even supermode. Reading the graph obtained at the figure 8 determines that the closer the guides are, the larger the effective index difference between the even and the odd supermodes at the point of intersection.

Indeed, at the point of intersection of the Ag and Si curve, the coupling constant is all the greater as the effective index difference between the even and odd supermodes is important, in agreement with the relation previous (1). A low interval Di, equivalent to an effective index difference of the large supermodes, is therefore preferable for increasing the efficiency of the coupling. However, the greater the effective index difference, the greater the strength or coupling power increases. A small interval is therefore preferable to increase the efficiency of the coupling.

Preferably, the interval Di separating the two guides has a minimum of 10 nm and a maximum of the order of the working or operating wavelength, that is to say function of the length of the wavelength. wave that one seeks to couple. Indeed, without offset, the separation interval could be zero and the guides in contact. Such contact between the two guides would generate a diffraction of the electromagnetic wave, and thus a loss of coupling efficiency. In addition, contacting the slot metal guide 2 with the silicon waveguide 1 could contaminate the optical properties by migration of the metal species in the silicon guide.

According to a third improvement, the partial overlap length L _r of the waveguide 1 by the slot metal guide 2 is an odd multiple of an effective coupling length Lc (equation (2)) between the two guides 1, 2 The effective coupling length Lc is defined as being the shortest length for which the power injected into the waveguide 1 is maximum in the slot metal guide 2. This power transfer in the slot metal guide 2 is period, said period being L _b = 2L _c , L _b being also called beat length.

Thus, the overlap length will preferably be equal to the coupling length or more generally to an odd multiple (2n + 1) * Lc for transmitting the electromagnetic wave from one guide to the other, with n a positive integer or no.

The use of a slotted metal guide 2 whose total length is equal to the overlap length which is an odd multiple of the coupling length Lc advantageously makes it possible to confine the TE component in the slot metal guide 2. , the TE component is found in the slotted guide 2 and the TM component continues to propagate in the waveguide 1.

The coupling length and coupling efficiency can also be optimized by a parametric study performed using numerical simulations using for example the finite-difference frequency-domain (FDTD) method. This numerical method is used to simulate the propagation of electromagnetic waves in structures. The procedure of the example uses a Gaussian beam generated in the silicon ribbon guide 1 and propagating along k. The working wavelength was chosen around 1550nm. According to the particular example of the silver slot guide 2 and the silicon guide 1, the coupling length Lc for which 75% of power is transmitted to the slot metal guide 2 is equal to 0.9 μm. It reflects a great difference in index of the two supermodes, that is to say a great coupling force which consequently allows a high compactness of the device. This characteristic also means that the phase shift between the two supermodes, in other words the coupling, will therefore be insensitive to the wavelength and the geometric dimensions of the guides (large manufacturing tolerance).

Of course, the various improvements mentioned above can be used in combination, their synergy makes it possible to obtain optimum efficiency of the coupling device

According to a particular embodiment illustrated in Figures 9 and 10 , the waveguide is formed by two elements 1a, 1b, preferably aligned along the same axis. The slotted metal guide 2 is arranged so as to cover partially two proximal portions 1a, 1b of said elements over lap lengths L _r1 and L _r2 . This makes it possible to produce a first coupling between the element 1a of the waveguide and the slot metal guide 2 in the direction of propagation k of the electromagnetic wave, and then a second coupling between the slotted metal guide 2 and the element 1b of the waveguide. According to one embodiment, only the TE components pass through the slot guide, the latter are the only ones to be coupled during the second coupling. Preferably, recovery of the lengths L and L _{r1 r2} of the waveguide by the slotted guide 2 at each element portion 1 a, 1 b are respectively equal to an odd multiple of the effective coupling length L _c . The methods for determining the effective coupling length L _c previously described apply to this embodiment.

When the waveguide is formed by two elements 1a, 1b, the two proximal portions may optionally be thinned to reduce the section of the waveguide so that the signal not coupled to the slot metal guide 2 is not reflected. by the end of the waveguide.

The coupling device, and its embodiments, may, in addition to allowing the coupling of the two guides, be used to produce optoelectronic components (detectors, emitters, nonlinear optical components) depending on the nature of the medium constituting In fact, such a coupling makes it possible to exacerbate the light-material interaction by confinement in the slot in order to make the devices more efficient (sensitivity, efficiency, non-linearity). The use of a slot metal guide 2 allows, depending on the application, to use the two metal elements 3a, 3b constituting it as electrodes for, for example, detecting or applying a voltage difference across said slot. The application of a voltage makes it possible, for example, to modulate the index of the material (for example SiOx) contained in the slot, that is to say between the two metallic elements 3a, 3b coplanar.

Another application may be the detection or the stimulated emission of sub-micronic objects such as nano-wires, nano-antennas, quantum boxes, etc. Indeed, TE components of the electromagnetic wave confined in the slot guide 2 can converge more easily to an object of a few nanometers placed in said slot. As before, the metal elements 3a, 3b of the slot guide 2 can form electrodes able to apply a stimulus or to measure values.

In general, the slot may also be filled with a material having particular optical properties such as non-linearity with materials of the SiOx type (silica loaded with silicon nanocrystals) or of the polymer type. The material may also have emission properties and be SiOx type (Silicon nanocrystal-loaded silica) doped with Erbium ions or the III-V family (AsGa or InP). The material may also have modulation properties and be of the BST or PZT type (ferroelectric materials which possess piezoelectric properties). In other words, the slot can be filled by a non-linear material, a material with emission properties or a material with modulation properties.

Alternatively, it is possible to fill the slot 4 with germanium or silica loaded with silicon nanocrystals and to use the metal elements 3a, 3b to apply a voltage to the material filling the slot to make it more or less transparent . This allows for example to form amplitude or phase modulation filters of the wave considered.

Another application may be the separation of TE and TM waves in two waveguides 1 a, 1 b. A device using the invention to achieve such function is illustrated at the figure 11 . Such a separation device comprises a first waveguide 1 coupled to a first slotted metal guide 2a. The overlap length L _r1 of the first slotted metal guide 2a on the first waveguide 1a is an odd multiple of the coupling length, thereby allowing the TE component to be confined in the first slotted metal guide 2a. In the particular case of figure 11 the waveguide 1a is continuous and the overlap length L _r1 is equal to the length Lf of the first slotted metal guide 2a. According to the axis k of displacement of the electromagnetic wave in the first waveguide 1a, only the TM component is present after the first metal slot guide 2a because the latter has not been coupled. The first slot guide 2a is coupled to a second slot guide 2b itself coupled to a second waveguide 1b. The overlap length L _r2 of the second waveguide by the slot guide is preferably equal to an odd multiple of the coupling length Lc to allow the transfer of the TE component into the second waveguide 1b. .

As an exemplary embodiment, the electromagnetic wave can be brought into the coupling device by an optical fiber either by the network method, or by the typing method.

In the array method, the waveguide 1 has a plurality of ribs on its surface, each rib preferably being perpendicular to the longitudinal axis of the waveguide. The optical fiber is oriented in a direction close to normal to the plane containing the ribs, that is to say the plane of the guide.

In the "tap" method the waveguide 1 comprises, for example, a divergent or convergent end (inverted tap) abutting with the optical fiber in the plane of the guides. Such an embodiment is described in the document "IEEE photonics technologiy letters" Volume 17 n ° 12 of Roelkens et al.

• former un guide d'onde 1 (figures 12 et 13),
• former un guide métallique à fente 2 (figures 15 et 16) dans un plan décalé par rapport au plan du guide d'onde 1 de sorte que le guide métallique à fente recouvre partiellement le guide d'onde 1 et que la fente 4 soit délimitée par deux éléments métalliques 3a, 3b coplanaires. De préférence, la fente 4 du guide à fente 2 est orientée selon un axe longitudinal A1 du guide d'onde. Comme indiqué précédemment, le décalage peut être réalisé par l'intermédiaire d'un matériau diélectrique.

A method of producing a coupling device illustrated in FIGS. Figures 12 to 16 comprises, on a substrate, the following successive steps:

• form a waveguide 1 ( Figures 12 and 13 )
• form a slotted metal guide 2 ( Figures 15 and 16 ) in a plane offset from the plane of the waveguide 1 so that the slotted metal guide partially covers the waveguide 1 and the slot 4 is delimited by two metal elements 3a, 3b coplanar. Preferably, the slot 4 of the slot guide 2 is oriented along a longitudinal axis A1 of the waveguide. As indicated above, the offset can be achieved via a dielectric material.

According to a development, after formation of the waveguide 1, the latter is encapsulated or partially covered, before the formation of the slotted guide 2, by a dielectric material 8 of low refractive index with respect to the refractive index of the waveguide 1, this dielectric material 8 is preferably silicon oxide ( figure 14 ). Low index means a material whose refractive index is between 1 and 2.5. This encapsulation allows in particular to shift the two guides from one another and to participate in the formation of the slotted metal guide 2.

Preferably, the coupling device can advantageously be produced as illustrated in FIG. figure 12 from a substrate 5 of the SOI type (silicon 6 on insulator 7). Thus, the step of forming the waveguide 1 is performed by partial or total etching of the upper layer 6 of silicon. By total etching is meant the etching of the upper silicon layer 6 to the insulating layer 7 to form the silicon waveguide 1. Partial etching makes it possible, for example, to form ribs for coupling to an optical fiber as described above.

• graver deux cavités 9a, 9b dans le matériau diélectrique 8 (figure 15) encapsulant ou recouvrant en partie le guide d'onde 1, à ce stade, les deux cavités 9a, 9b sont séparées par une paroi délimitant la largeur maximum et la hauteur maximum de la fente,
• remplir les deux cavités par du métal, de préférence de l'argent ou du cuivre (figure 16), l'épaisseur de métal déposée définissant la hauteur de la fente.

According to a variant, the formation of the slotted metal guide 2 is carried out by the following successive steps:

• to engrave two cavities 9a, 9b in the dielectric material 8 ( figure 15 encapsulating or partially covering the waveguide 1, at this stage, the two cavities 9a, 9b are separated by a wall delimiting the maximum width and the maximum height of the slot,
• fill the two cavities with metal, preferably silver or copper ( figure 16 ), the deposited metal thickness defining the height of the slot.

It is also possible to encapsulate the waveguide 1 and then to perform on the encapsulation a layer having particular optical properties as defined above (emission, modulation, non-linearity). Preferably, this layer will be germanium or silica loaded with nanocrystals. The cavities can then be made in this last layer and the slot will be automatically filled by the corresponding material.

Preferably, the slot guide 2 is made by the Damascene method, that is to say that the dielectric 8 encapsulating or partially covering the waveguide 1 is etched partially or totally above the waveguide 1 to make cavities of dimensions equal to the dimensions of the metal elements 3a, 3b and to define the slot 4 of the slot guide. The metal can then be deposited on the entire substrate before polishing comes, preferably, remove all the metal present outside said cavities 9a, 9b. In the case where there is total etching, the metal elements 3a, 3b are in contact with the waveguide 1. In the particular case described above, where the method comprises an additional layer having particular optical properties, the latter may to be engraved to the dielectric 8 and the metal elements 3a, 3b are then in contact with the dielectric 8.

Thus, the material in the slot, that is to say separating the two elements 3a, 3b forming the slot metal guide 2 may be different from that which encapsulates the waveguide 1 or partially covers the waveguide 1. The material can also be removed after formation of the metal elements 3a, 3b to leave the slot free.

The Damascene method has the advantage of allowing the production of a slot 4 between the two metallic elements 3a, 3b, preferably of copper and / or silver, the edges of which are more abrupt and less rough than those obtained by direct etching. metals. Preferably, the metal used to form the metal elements 3a, 3b is a metal having a low refractive index. This improves the operation of the slotted metal guide by reducing the loss of dissipation that can be caused by the roughness of the slot. This method also allows the use of materials whose direct etching is not easily reproducible such as copper.

According to one embodiment, the method comprises between the step of etching the cavities 9a, 9b and the cavity filling step 9a, 9b a chemical etching step at said cavities 9a, 9b to allow the thinning of the wall defining the dimensions of the slot.

It is thus possible to make a slot whose width can be between 10nm and 100nm thanks to optical lithography (DUV).

According to a particular embodiment, after the step of filling the cavities to form the slotted guide, a step of selective etching of the material located between the two metallic elements is carried out in order to allow the filling of the slot by another material different from that encapsulating the waveguide 1.

One of the applications of this embodiment is particle detection. These can be trapped in the vent slot, preferably by a microfluidic system, to modify the propagation constant of the electromagnetic wave in said slot guide and thus the coupling efficiency to the guide. wave. The variation of the efficiency then makes it possible to detect certain particles.

In order to protect the slot guide 2, it is possible to encapsulate it with a dielectric with a low refractive index, for example silicon oxide. Such an index of refraction can be between 1 and 2.5

The coupling device is preferably made based on non-contaminating metals (copper, aluminum, etc.) to be able to use the same manufacturing means as those used in the fields of microelectronics and photonics on silicon.

The coupling device described above can serve as an elementary brick in an optical circuit and allows optical interconnections to be made at very small distances. It has the advantage of a very good efficiency and a high compactness.

According to one approach, it will be sought to use the coupling device to form an optical and electrical coupler of an object, for example a molecule preferably of nanometric dimensions. Indeed, the use of a thin-film slot guide, for example of the order of 50 nm thick, makes it possible to address objects of nanometric dimensions. Such a coupler makes it possible in particular to link the electrical transport properties of the object to its optical properties. Application examples can be molecular spectroscopy to study the electronic states of an unknown object. For a known object, it is possible to optically modulate the conductance of the object, and thus to produce a molecular transistor controlled by an optical gate. The invention may also be useful for coupling the luminescence of an object to a silicon waveguide.

One solution to this approach is to combine an optical coupling as described using a waveguide, and a slot guide combined with an electrical coupling.

As illustrated in Figures 17 to 19 an optical and electrical coupler comprises a coupling device as described above, according to its various embodiments and variants, and an electric coupling element provided with two electrodes arranged at the slot of the slot guide 2. The object is preferably in contact with at least one of the electrodes. Of course, the slot guide 2 is optically coupled to the waveguide 1. The object 11 may also be in contact with the two electrodes, in both cases the object in electrical contact with the electrode or electrodes may then be addressed electrically.

The electrodes preferably comprise two microtips 10a, 10b, preferably metal, placed facing one another, one of the microtips 10a being oriented towards the other microtip 10b. The microtips are used to address objects of smaller dimensions than the slot. For objects larger than the slot (nanowires, graphene nanotubes, etc.) the electrodes may have any shape.

In fact, the acute end of a microtip 10a, 10b is proximal to the sharp end of the other microtip. The object 11 may be in electrical contact with one of the microtips 10a, 10b, at its acute end, or be electrically clamped by the two microtips 10a, 10b ( Figures 17 and 18 ). Of more generally, the object 11 is in electrical contact with at least one of the electrodes.

According to a first embodiment of the coupler illustrated in FIG. figure 17 the electrodes are formed by the metallic elements 3a, 3b of the slotted guide 2, and each of the metal elements 3a, 3b comprises at the slot separating them a microtip, one of the microtips 10a, 10b being oriented towards the another microtip. The microtips 10a, 10b are useful for coupling nanoscale molecules.

In fact, the two metal elements 3a, 3b may comprise two edges facing each other and delimiting the slot, a microtip may be made on one of the edges which locally forms a projection towards a substantially identical projection of the other edge. In other words, the microtips 10a, 10b are arranged in the slot, each of the metal elements 3a, 3b then comprises in the slot separating them a microtip 10a, 10b, one of the microtips being oriented towards the other microtip. In this case, the slot guide 2 makes it possible on the one hand to optically couple the wave to the object 11, and on the other hand to electrically couple the object 11, for example by connecting the two metal elements 3a, 3b to a measurement and / or polarization electronics.

According to a second embodiment illustrated in Figures 18 to 19 , the metal elements 3a, 3b of the slot guide 2 are located between the waveguide 1 and the electrodes, preferably having microtips 10a, 10b. Unlike the above mode, the electrodes are then distinct from the metal elements 3a, 3b but remain at the level, that is to say near the slot. Said electrodes may be between 0 nm and 500 nm of the metal elements 3 a, 3 b, and generally up to half the incident wavelength, and preferably at a minimum distance greater than the width of the slot. In other words, the electrodes are located in a plane shifted from the plane of the metal elements 3a, 3b, and are preferably arranged above the slot 4. In fact, in this second embodiment, the electrodes are distinct from the metal elements 3a, 3b. Preferably, in the second embodiment, the electrodes are separated from the metallic elements 3a, 3b by a layer of a dielectric material, for example silicon oxide, of low index relative to the waveguide 1, typically the refractive index of the dielectric material is between 1 and 2.5. The second embodiment has the advantage of not interfering with the optical coupling of the waveguide 1 with the slot guide 2 with respect to the first embodiment of the coupler.

The thickness of the electrodes of the second embodiment in a direction perpendicular to the plane of the waveguide may be between a few nanometers at a zone intended for the electrical coupling of objects of nanometric size (molecules), and a hundred or so nanometers at a contact recovery zone to facilitate said resumption of electrical contact. In other words, each electrode may comprise an electrical coupling zone (addressing the object) whose thickness is less than a contact recovery zone of the electrodes. Of course, in the first embodiment of the coupler, the thickness of the electrodes is equal to the thickness of the metal elements 3a, 3b. In fact, a thin thickness at the electrodes makes it possible to contact an object of nanometric size (for example a molecule). For large objects such as nanowires, nanotubes, graphene, the more the contact at the coupling zone will be thick the better it will be.

In both embodiments of the coupler comprising the microtips 10a, 10b, the latter are preferably located at the median of the slot, said median being perpendicular to the longitudinal axis of the slot. This allows in particular to reduce the surface of the electrodes contained in or near the slot to disturb the propagation of the light signal to a minimum. The spacing between the two electrodes (for example via the microtips) is preferably less than 200 nm. For larger objects than the slot, the electrodes may be offset from each other, and may be spaced apart larger than the width of the slot.

In order to achieve a spacing between the microtips of less than 30 nm, these can be obtained by forming electrodes in electrical contact separated by a relatively thin bridge (for example about ten nanometers thick, and about 50 nm x 100 nm side dimensions) . Then these electrodes can be polarized in current so as to break the bridge, and space the two electrodes by a nanoscale gap. It can also be envisaged to apply mechanical stresses to the bridge to break it and delimit the microtips. For spacings greater than 30nm, lithography / engraving techniques will be used.

Preferably, the microtips 10a, 10b and the metal elements 3a, 3b of the slot guide 2 are made of different materials to ensure good ohmic contact of the points with the object. In general, the microtips are formed by metal multilayers such as for example Al or Au.

The object 11 may be an object of nanometric size such as a nanotube, graphene, or a quantum dot whose growth and then deposition can be made so as to bring said box into contact with the electrodes.

An electrical contact between the object and the electrodes can be obtained by electromigration when the object is a molecule, or one or more metal particles (gold balls for example).

To summarize, if the object is of nanometric size, for example a molecule, electrodes comprising microtips will preferably be used, and for larger objects, such as nanowires, carbon nanotubes or graphene, conventional electrodes may be used. When used, the electrodes will then preferably be arranged so as not to impede the propagation of the light signal in the slot guide. The metal elements alone can also serve as electrodes.

The coupler offers a high optical coupling efficiency towards a nano-sized object, because the wave confined in the slit guide slot has a concentration one hundred times higher than the concentration that can be found in the prior art. .

Claims (18)

1. Device for co-directional coupling an electromagnetic wave including:

   - a waveguide (1),

   - a slit metal guide (2),

   the slit metal guide (2) being formed by two metal elements (3a, 3b) which are coplanar and spaced out from one another so as to delimit the slit, and the slit metal guide (2) being arranged in a plane (P2) offset from the plane (P1) of the waveguide (1) and partially covering said waveguide (1), said waveguide (1) and the slit guide (2) being maintained at a distance from one another by a dielectric.
2. Device according to claim 1, characterized in that the waveguide (1) and the slit metal guide (2) are separated by an interval (Di).
3. Device according to anyone of the claims 1 and 2, wherein characterized in that the waveguide (1) and the slit metal guide (2) have effective indices whose ratio lies between 0.8 and 1.2.
4. Device according to anyone of the claims 1 to 3, characterized in that the partial covering length (L_r) of the waveguide (1) by the slit metal guide (2) is an odd multiple of a effective coupling length (Lc) between the two guides.
5. Device according to anyone of the claims 1 to 4, characterized in that, the waveguide being formed by two elements (1 a, 1 b), said slit metal guide (2) is arranged so as to partially cover two proximal portions of said elements (1 a, 1 b).
6. Device according to claim 5, characterized in that the partial covering length (L_r1, L_r2) of the waveguide by the slit metal guide (2) at each portion is an odd multiple of a effective coupling length (Lc).
7. Device according to anyone of the claims 1 to 6, characterized in the metal elements (3a, 3b) form electrodes.
8. Device according to claim 7, characterized in that the slit (4) is filled with germanium.
9. Device according to claim 7, characterized in that the slit is filled with a nonlinear material or a material with emission properties or a material of modulation properties.
10. Device according to claim 7, characterized in that each of the metal elements (3a, 3b) comprises at the slit a microtip, one of the microtips being oriented towards the other microtip.
11. Method for manufacturing a coupling device between a waveguide (1) and a slit metal guide (2), characterized in that it comprises on a substrate (5) the following successive steps:

    - forming the waveguide (1),

    - forming a dielectric material layer (8),

    - forming the slit metal guide (2) so as to be co-directional with the waveguide (1) in a plane offset from the plane of the waveguide (1) so that the slit (4) is delimited by two coplanar metal elements (3a, 3b) and the slit metal guide (2) partially covers the waveguide (1), the waveguide (1) and the slit metal guide (2) being separated from each other by the dielectric material layer (8).
12. Method according to claim 11, characterized in that the waveguide (1) is encapsulated, or partly covered, by a dielectric material (8) whose refractive index lies between 1 and 2.5 before the formation of the slit guide (2).
13. Method according to anyone of the claims 11 and 12, characterized in that, as the substrate (5) is a substrate provided with a silicon (6) on insulator layer (7), the waveguide (1) is a silicon guide delimited by etching the silicon layer (6) down to the insulating material layer (7).
14. Method according to claim 12, characterized in that the formation of the slit metal guide (2) is carried out by the following successive steps:

    - etching two cavities (9a, 9b) into the dielectric material (8) encapsulating or partly covering the waveguide (1),

    - filling the two cavities (9a, 9b) with metal.
15. Method according to the claim 14, characterized in that, between the step of etching the cavities (9a, 9b) and the step of filling the cavities (9a, 9b), it comprises a chemical etching step at said cavities (9a, 9b).
16. Electric and optical coupler of an object, characterized in that it comprises coupler comprises a coupling device according to claim 1, and an electric coupling element provided with two electrodes arranged at the slit of the slit guide (2), said object being in contact with at least one of the electrodes.
17. Coupler according to claim 16, characterized in that the electrodes are formed by the metal elements (3a, 3b) of the slit guide (2), and in that each metal element (3a, 3b) comprises in the slit thereb0etween a microtip (10a, 10b), one of the microtips (10a, 10b) being oriented towards the other microtip (10a, 10b).
18. Coupler according to claim 16, characterized in that the metal elements (3a, 3b) of the slit guide are located between the waveguide (1) and the electrodes.

Images